C. L. Speyers — Molecular Weights of Liquids. 427 



Art. XLI. — Molecular Weights of Liquids, ivith a few 

 Words about Association ; by C. L. Speyers. 



For some years past, evidence has been advanced* to show 

 that 



n _ p-p' 



N ~ p' [ ' 



agrees with experiment far better than 



does. While (1) is as true for concentrated solutions as for 

 dilute ones, (2) fails both in theory and in fact for concen- 

 trated solutions, though of course good for dilute ones. 



A theoretical foundation was sought for (1) in the notion 

 that a volatile liquid produces those molecules which pass off 

 as vapor in the body of the liquid instead of merely on the 

 surface, a notion which J. Traubef is advancing as one of the 

 results of his study of critical phenomena, and that the vapor 

 tension of a liquid is the result of such differentiation in the 

 liquid. 



Moreover, on combining (1) with the second law of thermo- 

 dynamics we get equations for the molecular rise in boiling 

 point and depression in freezing point, for the molecular heat 

 of vaporization and of freezing, and all these independently of 

 the osmotic theory. 



Equation (1) has also been advanced tentatively by S. Young 

 and E. C. Forteylj: but with some reservations in regard to 

 changes in density. The new mixtures that they used were ethyl 

 propionate with ethyl acetate, toluene with ethyl benzene, n 

 hexane with n octane and toluene with benzene. Slight dif- 

 ferences were found between observed pressures and those cal- 

 culated by (1), but no account was taken of a possible association 

 of the vapor of the solvent and so these differences lose some- 

 thing of their significance. In their paper, the writers correct 

 a mistake made by me. They point out that mixtures of non- 

 associated liquids may have minimum boiling points. I stated 

 they could not, and in this respect was wrong, having over- 

 looked the fact that Ap 2 /At could be smaller than Ap x /At at 

 one temperature while greater at another, both temperatures 

 lying within the range of boiling points of the mixtures. 



* Journ. Phys. Chem., ii, 347, 362(1898); Journ. Am. Chem. Soc, xxi, 

 282 (1899) ; this Journal, ix, 341 (1900) ; xiii, 213 (1902). 

 f Drade's Ann., viii, 267 (1902). 

 X Journ. Lond. Chem. Soc, lxxxiii, 45, 68 (1903). 



Am. Jour. Sci. — Fourth Series, Vol. XVII, JS r o. 102. — June, 1904. 

 29 



